Monkey's Memory Cells Caught In The Act Of Learning

The researchers measured activity of individual neurons in the hippocampus as two monkeys formed new associative memories while playing a computer game. The monkey had to gaze at one of four identical targets superimposed on a complex visual scene to win a juice reward. For example, white dots appeared at four locations over a picture of a gorilla in a forest. The animals learned, through trial and error, which location was associated with the reward for each new scene. Eye-tracking technology detected if the animal was fixated on the correct dot and the reward was dispensed. Monkeys learned 2-4 new associations during each recording session. It took about a dozen trials to establish the memory for each scene.

Among 89 neurons that responded to the scenes, the researchers discovered a subset of 25 "changing cells" whose increase or decrease in activity paralleled learning, signaling that they were involved in the learning process. The changes in neural activity took place before, during or immediately after learning, suggesting that there is a gradual recruitment of a network of hippocampal neurons associated with the formation of new associative memories. By relating the timing of a neuron's changes to the animal's behavioral performance, the researchers were able to precisely pinpoint its role in learning. Some neurons changed only in response to one of several scenes, indicating a high level of specialization for particular stimuli.

Some of the changing cells showed sustained activity after the trials were over, while others returned to baseline. The researchers hypothesize that these "sustained changing cells" may not only participate in memory formation, but also in storage of the associations in long-term memory.

"Changing cells" were not detected until now because studies of long-term memory have typically not begun monitoring neurons in animals until after they had already been trained on experimental tasks for months, according to Suzuki. Also, the signals are hard to detect if the learning task is too difficult or too easy, resulting in only shallow increments of learning.

"This study provides a direct demonstration of learning-related neural plasticity in the hippocampus," notes Suzuki.

In addition to Dr. Suzuki, Dr. Sylvia Wirth and Marianna Yanike, NYU, conducted the recording experiments. Drs. Loren Frank, Anne Smith and Emery Brown, Harvard Medical School, developed algorithms used to analyze the neural and behavioral data.

The research was also supported by grants from the McKnight Foundation and the John Merck Fund.

NIMH and NIDA are part of the National Institutes of Health (NIH), the Federal Government's primary agency for biomedical and behavioral research. NIH is a component of the U.S. Department of Health and Human Services.